Design and optimization of high refractive index proteins for deep tissue imaging in living biological systems - PROJECT ABSTRACT Light microscopy is uniquely advantageous in biomedical research due to its speed, resolution, and noninvasiveness, but it is unable to visualize dynamic processes occurring deep within tissues, hindering our ability to fully understand complex, multilayered biological information in a native context. Live imaging of intact tissue is universally hindered by light scattering across all forms of optical microscopy and tissue types. Despite decades of work toward improving imaging depth with new hardware designs, computational methods, and optical probes, scattering remains a broad limitation in biomedical studies because these methods have not addressed the core problem of refractive index (RI) mismatching between the aqueous and lipid fractions of cells. Traditional methods for reducing light scattering such as immersing tissues in high-RI solutions or using clearing agents, are limited to fixed or ex vivo samples and cannot be applied to living tissues without compromising viability. To overcome this technology roadblock, this proposal details an innovative and iterative strategy to combine AI-driven design of synthetic high RI proteins (Aim 1), to systematically characterize the properties of ~50-100 candidates in cells using a custom, scalable pipeline (Aim 2), and by advancing ~15 high priority candidates for testing safe, genetic delivery in a range of 3D living tissue systems (Aim 3). Our stringent and iterative strategy builds on our compelling prior and preliminary work as well as multi-disciplinary expertise in AI-driven synthetic protein design, gene expression and delivery to tissues, and inventing optical imaging methods. Moreover, genetic encoding offers an advantage over prior methods because it circumvents challenges in diffusion, intracellular access, and allows tunable spatiotemporal expression of RI proteins making it a high value target to enhance the effect of or replace chemical methods. The approaches are entirely unique to this proposal and the synergistic effort of our team is critical to ultimately realize the potential of a tunable, genetically encoded tool for achieving live tissue transparency and access to greater imaging depths in bulk tissues. The data, insights, live imaging assays, gene constructs, tissue culture protocols, computational design tools and image analysis pipelines generated during this work will be freely shared to accelerate a new field of genetic RI manipulation toward development of safe, effective, and accessible research tools. Thus, our work has the potential to dramatically improve live imaging speed and resolution over large tissue volumes that will be of broad utility and impact across biology and medicine.
